Open AccessArticle

Predicting Melt Pond Fraction on Landfast Snow Covered First Year Sea Ice from Winter C-Band SAR Backscatter Utilizing Linear, Polarimetric and Texture Parameters

Saroat Ramjan

¹,

Torsten Geldsetzer

Randall Scharien

and

John Yackel

^1,*

Cryosphere Climate Research Group, Department of Geography, University of Calgary, Calgary, AB T2N 1N4, Canada

Department of Geography, University of Victoria, Victoria, BC V8W 2Y2, Canada

Author to whom correspondence should be addressed.

Remote Sens. 2018, 10(10), 1603; https://doi.org/10.3390/rs10101603

Submission received: 28 June 2018 / Revised: 9 September 2018 / Accepted: 2 October 2018 / Published: 9 October 2018

(This article belongs to the Special Issue Cryospheric Remote Sensing II)

Download

Browse Figures

Versions Notes

Abstract

Early-summer melt pond fraction is predicted using late-winter C-band backscatter of snow-covered first-year sea ice. Aerial photographs were acquired during an early-summer 2012 field campaign in Resolute Passage, Nunavut, Canada, on smooth first-year sea ice to estimate the melt pond fraction. RADARSAT-2 Synthetic Aperture Radar (SAR) data were acquired over the study area in late winter prior to melt onset. Correlations between the melt pond fractions and late-winter linear and polarimetric SAR parameters and texture measures derived from the SAR parameters are utilized to develop multivariate regression models that predict melt pond fractions. The results demonstrate substantial capability of the regression models to predict melt pond fractions for all SAR incidence angle ranges. The combination of the most significant linear, polarimetric and texture parameters provide the best model at far-range incidence angles, with an

R^{2}

of 0.62 and a pond fraction RMSE of 0.09. Near- and mid- range incidence angle models provide

R^{2}

values of 0.57 and 0.61, respectively, with an RMSE of 0.11. The strength of the regression models improves when SAR parameters are combined with texture parameters. These predictions also serve as a proxy to estimate snow thickness distributions during late winter as higher pond fractions evolve from thinner snow cover.

Keywords:

melt pond fraction; snow; SAR; polarimetric parameters; GLCM texture

1. Introduction

The spatial distribution, evolution, and extent of melt ponds on sea ice are functions of snow thickness and snow redistribution processes through the preceding winter and spring, ice surface roughness, and ice type [1,2,3,4]. For spatially varying snow covers on sea ice, thicker snow takes longer to melt than thinner snow under consistent atmospheric forcing [1,5,6]. Prior studies advocate that a thinner snow cover leads to dominant surface flooding (i.e., larger areal melt pond fraction,

f_{p}

), whereas a thick winter snow cover leads to a greater fraction of snow patches and thus less surface flooding (i.e., smaller

f_{p}

) [7,8]. Thicker snow covers generally accumulate on comparatively rough first-year sea ice (FYI), as the uneven surface topography effectively captures wind-blown snow [9]. The relatively smooth topography of FYI causes thinner snow covers, with consistent drifting of snow following the wind direction during depositional storm events [10]. Therefore, smooth FYI, with an associated thin snow cover, leads to the horizontal spreading of ponds over a larger area, with

f_{p}

as high as 80% [3,4,11]. Since the formation of FYI melt ponds is influenced by winter snow thickness,

f_{p}

could provide a proxy for winter ice surface roughness and snow thickness.

After melt onset, melt ponds gradually develop over four successive stages. These stages have been previously identified as: (1) topographic control, (2) hydrostatic balance, (3) ice freeboard control and (4) fall freeze up or ice break up [6,8,12]. During stage 1, melt ponds start to form and eventually fill topographic depressions with minimal drainage through the ice. At this stage, the structure of the melt ponds is very unstable and is actually controlled by the pre-melt ice topography [1]. Stage 2 begins when melt ponds have the same production and drainage rates, and lateral melt water flows towards macroscopic fractures and seal holes promote the formation of melt pond networks by interconnecting melt ponds. Consequently, the geometric structure of melt pond covers becomes prominent during this stage [1,12]. Stage 3 is characterized by increased vertical drainage and decreased horizontal discharge to macroscopic fractures. Surface topography of melting and decaying ice cover controls the melt pond coverage in this stage [6]. Finally, during stage 4, complete decay or disintegration of FYI, or fall freeze up of multiyear sea ice begins.

Estimating melt pond fraction over large areas requires satellite-based methods. Optical approaches [13,14,15,16,17,18] are hampered by cloud cover and are thus limited by the accuracy of cloud identification and screening techniques. They must also contend with regional differences in spectral properties [18], and issues related to sub-pixel variation. The root-mean-squared error (RMSE) for optical approaches is 0.08 to 0.16. Approaches using passive microwave data [19] are limited by the sensor resolution, especially in narrow channels where land contamination is likely. They are also only applicable to very high sea ice concentrations; even small areas of open water would dominate the signal.

Approaches using active microwave data have provided varying results. Barber and Yackel [20] report a relationship between ERS-1 backscatter and

f_{p}

, but note a limitation related to wind roughening of melt ponds. Mäkynen et al. [21] report poor skill in connecting changes in

f_{p}

with backscatter magnitude and texture using ENVISAT WSM images of sea ice during the ponding stages. Scharien et al. [22] report on the usefulness of co-polarization and cross-polarization ratios in estimating

f_{p}

during ponding using RADARSAT-2 quad-pol data; their RMSE values range from 0.05 to 0.43. Using TerraSAR-X, Fors et al. [23] also find the co-polarization ratio useful during the ponding stage, at intermediate wind speeds, with an RMSE is 0.4; they also evaluate four polarimetric parameters and find reasonable correlations. To evaluate prediction of

f_{p}

from winter images, Scharien et al. [24,25] use ENVISAT-ASAR and Sentinel-1 co- and cross-polarized backscatter and texture measures thereof. They report overall RMSE values of 0.08 and 0.09, but acknowledge model underestimation of

f_{p}

for smooth FYI related to sensor noise floors.

Estimating melt pond fraction over smaller areas and for validation of the techniques above rely on in-situ observations [11,22], or low-level aerial photographs [26,27]. A recent study using low-level aerial photographs showed that, due to their fine spatial and temporal resolutions, they are an ideal data source for estimating

f_{p}

since image classification accuracies are as high as 97% [22].

Using winter imagery to predict

f_{p}

implies an association of sea ice roughness and/or its snow cover variability with subsequent

f_{p}

. Comprehensive studies on the degree of winter snow thickness spatial variability, as well as relationships with

f_{p}

, are essential to better understand radiative transfer processes leading to snow and sea ice ablation rates during the spring melt transition. However, studying snow thickness variability on sea ice is complicated. The accumulation and redistribution of snow on FYI exhibits high spatio-temporal variability [9], and logistical challenges, resulting in a shortage of in-situ data [10]. Snow thickness distributions have been obtained in-situ [9,28,29,30], using laboratory-based procedures [31,32], and using remote sensing methods based on passive microwave data [33,34,35], frequency-modulated continuous-wave airborne radar returns [36,37] and Synthetic Aperture Radar (SAR) data [7,38,39], and using a combination of active microwave scatterometer and optical data [40].

The use of SAR backscatter must also contend with variability in backscatter with incidence angle (

θ

). Kim et al. [41] identified the

θ

effect on backscatter intensity, as it strongly influences surface and volume scattering processes. Microwaves are dominated by surface scattering at lower

θ

(<30°); whereas at higher

θ

(>30°), volume scattering dominates over surface scattering [42,43].

The advent of polarimetric SAR data provide additional parameters for characterizing sea ice surfaces. Polarimetric parameters such as the Co-polarized phase difference, Co-polarized correlation coefficient, Entropy, Anisotropy, and Alpha angle can be calculated from second-order derivatives of the scattering matrix (i.e., covariance and coherency) to obtain enhanced information about both surface and volume scattering mechanisms of snow and sea ice [39,44,45,46,47].

In addition to linear and polarimetric parameters, SAR image texture also provides valuable structural information, leading to higher winter sea ice classification accuracy when compared to backscatter intensity alone [48,49,50,51]. Second-order texture measures, derived from the gray-level co-occurrence matrix (GLCM) introduced by Haralick et al. [52], were evaluated for sea ice classification [49], showing that the combination of

σ^{0}

(tone/grey-level) and texture measures (spatial dependence in tone) gives better results than using texture measures alone. GLCM texture is the most commonly used texture analysis technique; as it takes into account the spatial organization of grey tones within a moving window and offers a second-order statistical technique for extracting texture features [49,52]. Image texture of snow covered sea ice is a function of its near surface characteristics e.g., snow properties, snow thickness variability, and ice roughness [53]. Several studies successfully demonstrated the potential of GLCM texture measures, calculated from SAR images, to improve classification/segmentation of snow covered sea ice [49,51,53,54].

Given that the most consistent skill in estimating

f_{p}

is reported by winter prediction of

f_{p}

[24], we build on the winter prediction technique in this paper. We maintain the use of texture parameters as in Mäkynen et al. [21] and Scharien et al. [24] and expand on the analysis by evaluating the contributions of polarimetric parameters as per Fors et al. [23], and we assess texture measures of the polarimetric parameters. Furthermore, given the relationship of winter snow thickness and subsequent

f_{p}

, we attempt an initial inversion of

f_{p}

to estimate snow thickness.

Objectives

The primary objective of this study is to predict early summer

f_{p}

from winter C-band SAR polarimetric backscatter. A secondary objective is to relate the predicted

f_{p}

to the spatial distribution of snow thickness. Meeting these objectives will aid in understanding the relationship between snow thickness variability on FYI during winter and its evolution into a melt pond icescape in early summer. Since the SAR backscatter coefficient

σ^{0}

is largely a function of

θ

, this study will also examine the

θ

dependency on backscatter and its relationship with

f_{p}

. The following research questions towards these objectives are stated:

What are the relationships between winter C-band RADARSAT-2 (RS-2) SAR backscatter (linear and polarimetric parameters and texture parameters) and $f_{p}$ ?
How does the relationship between RS-2 SAR backscatter and $f_{p}$ change with $θ$ ?
Can we predict $f_{p}$ based on linear, polarimetric and texture parameters, and can predicted $f_{p}$ be used to infer the late winter snow thickness variability?

2. Methodology

2.1. Study Area

The study area is located in Resolute Passage, Nunavut, Canada and comprises of a large area of landfast FYI. The study area was imaged by the RADARSAT-2 SAR satellite during late-winter and by aerial photography during the melt pond season (Figure 1). Analyzing the melt pond evolution stages, we consider that stage 2 is a suitable time to examine aerial images of

f_{p}

and its relationship with late winter snow thickness. In-situ observations were carried out from late-winter through the melt pond season at a field site ~4 km southwest of Sheringham Point, Cornwallis Island. Fieldwork was conducted in two coordinated sampling periods (Legs). Leg 1 field work was conducted during the late-winter/early-spring period (8 to 29 May 2012), while Leg 2 was during late-spring/summer (4 June to 2 July 2012). Melt-onset was observed on 1 June 2012. Throughout Leg 1, the snow thickness range in the study region was 1.5 cm to 40 cm. Snow depths from snow pits and 5 transect lines 100 to 150 m long were collected. One snow depth measurement was acquired from each pit (a total of 102), and 565 snow depth measurements were sampled from all transects at 1 m spatial intervals. For this study, we used 202 samples of snow depth, including 102 measurements from the snow pits, and 100 measurements randomly selected from the snow transects to minimize spatial autocorrelation (Figure 2a) [9]. SAR acquisitions and snow samples were obtained at air temperatures between −20 °C and −3 °C (Figure 2b). An autonomous 2.75 m tall micro-meteorological station was installed on the sea ice from 18 May to 25 June 2012 to monitor air temperature, wind speed, and direction at 1-min intervals.

2.2. Data

2.2.1. Aerial Surveys

In early summer, the field site was initially characterized by a very shallow and wet snow cover overlying a rough superimposed ice layer. Melt ponds started to form on 10 June 2012, were fully formed on 15 June reaching at a maximum pond fraction of 78%, and remained above 50% until 25 June [12]. To quantify the evolution of melt ponds, their morphology, and

f_{p}

at a regional scale, four aerial photography surveys were flown between 13 and 29 June 2012 via fixed-wing Twin Otter aircraft. A Canon G10 camera attached to an opening in the hindmost of the plane acquired digital optical images at nadir viewing angle, with a sampling rate resulting in 0% to ~10% image overlap. Data from the second survey flight on 22 June 2012, 01:43 UTC (Figure 1) are used in this study, while the melt ponds were fully formed and the geometric structures of the melt ponds were established. The survey was flown at an altitude of 610 m in a grid pattern that also coincided with the field study site. The image data has a swath width of 750 m and pixel size of 0.22 m [22]. At the time of acquisition, the surface air temperature was approximately 2.6 °C, and the melt ponds were liquid and moderately wind roughened over the entire region; the wind speed was 3.2 m/s at the on-ice meteorological station and 5.3 m/s at the Resolute weather station.

2.2.2. RADARSAT-2 Data

RS-2 SAR scenes were acquired in fine quad-pol mode over the aerial flightpath at near-range (18°–29°) (NR), mid-range (30°–39°) (MR) and far-range (40°–49°) (FR)

θ

. Each scene consists of a polarimetric (HH + VV + HV + VH and inter-channel phase information) data set, with a nominal 5.2 m × 7.7 m resolution in range and azimuth, respectively, and covers a 25 km × 25 km area (MDA, 2014). This study uses seven RS-2 scenes acquired during Leg 1, that encompass the majority of aerial photographs of melt ponded sea ice acquired during Leg 2 (Table 1).

2.3. Methods

2.3.1. Estimation of Melt Pond Fraction from Aerial Photographs

f_{p}

was calculated from aerial photographs using a decision-tree classification approach [22]. Melt ponds are distinguished from adjacent ice/snow patches because the albedo of ponds is significantly lower than the snow and/or ice. The resulting classification (e.g., Figure 3a) has an error of 0.03 [22]. The distribution of

f_{p}

for the 22 June 2012 flightpath is presented in Figure 3b.

2.3.2. RADARSAT-2 Data Processing

RS-2 data were calibrated, and speckle filtered using a 5 × 5 boxcar filter to reduce the speckle effect while preserving image information [55,56]. Four linear and five polarimetric parameters were derived (Table 2). Polarimetric parameters were derived from covariance (1) and coherency matrices (2), as applicable [57,58]. Eight GLCM texture parameters (Table 3) were calculated for each linear and polarimetric parameter (Table 2) using a 5 × 5 moving window. Therefore, a total of 81 parameters (image bands) were generated. All parameter images were projected to a common geographic coordinate system using nearest neighbor intensity interpolation, at 5 m pixel spacing.

[C] = | \begin{matrix} 〈 {| S_{H H} |}^{2} 〉 & \sqrt{2} 〈 S_{H H} S^{*}_{H V} 〉 & 〈 S_{H H} S^{*}_{V V} 〉 \\ \sqrt{2} 〈 S_{H V} S^{*}_{H H} 〉 & 2 〈 {| S_{H V} |}^{2} 〉 & \sqrt{2} 〈 S_{H V} S_{V V}^{*} 〉 \\ 〈 S_{V V} S^{*}_{H H} 〉 & \sqrt{2} 〈 S_{V V} S^{*}_{H V} 〉 & 〈 {| S_{V V} |}^{2} 〉 \end{matrix} |

(1)

[T] = | \begin{matrix} 〈 {| (S_{H H} + S_{V V}) |}^{2} 〉 & 〈 (S_{H H} + S_{V V}) {(S_{H H} - S_{V V})}^{*} 〉 & 2 〈 (S_{H H} + S_{V V}) S^{*}_{H V} 〉 \\ 〈 (S_{H H} - S_{V V}) {(S_{H H} + S_{V V})}^{*} 〉 & 2 〈 {| (S_{H H} - S_{V V}) |}^{2} 〉 & 2 〈 (S_{H H} - S_{V V}) S^{*}_{H V} 〉 \\ 2 〈 S_{H V} {(S_{H H} + S_{V V})}^{*} 〉 & 2 〈 S_{H V} {(S_{H H} - S_{V V})}^{*} 〉 & 4 〈 {| S_{H V} |}^{2} 〉 \end{matrix} |

(2)

2.3.3. Image Sampling

Polygons corresponding to each 750 m by 750 m aerial photograph were used to extract spatially coincident SAR backscatter statistics from the seven RS-2 SAR scenes. The number of samples varied according to the portions of the RS-2 scenes that overlapped aerial photographs (Table 1). The total number of samples is 650, distributed across three different

θ

ranges.

Sample means of the nine linear and polarimetric parameters and eight GLCM parameters were calculated. Therefore, each of the 650 samples contains 81 associated mean values, corresponding to the nine linear and polarimetric parameters and their 72 derived texture parameters. Given that many adjacent aerial photographs overlap each other by up to approximately 10%, not all samples are entirely independent. However, the sample means and texture parameters represent the polygons as a whole, with no single-sample redundancy. Samples were divided into two equally sized training and test sets using a random sampling approach. A set of 325 training samples were used for model development, and 325 test samples were used to validate the derived models.

2.3.4. Regression Model Development

Multivariate linear regression was employed to generate combined backscatter, polarimetric parameter, and texture-based models to predict

f_{p}

. Spearman’s rank correlation analysis was used to assess correlation between parameters and with

f_{p}

while accounting for any non-normal distributions. A coefficient threshold of 0.7 was used to identify multicollinearity between parameters. Correlated parameter pairs were assessed against

f_{p}

, and the parameter with the higher correlation with

f_{p}

was kept for regression model development. Model selection was then accomplished using forward stepwise regression analysis with an entry p-value of 0.05. These analyses were performed separately for the three

θ

ranges (NR, MR and FR). Therefore, a different model is derived for each of the three

θ

ranges.

To identify any misspecifications in the multivariate regression models, and to support the accuracy values obtained with multivariate regression, a random forest regression [66] was also performed, using the random Forest package in the statistical package-R. The number of parameters used at each split was set to the number of parameters divided by 3, and 500 trees were used. The reduced parameter set, following removals due to correlation, was used for the random forest regression as well, since multicollinearity has been found to be detrimental [67].

Model accuracy was evaluated via RMSE analysis of predicted

f_{p}

values versus observed values in the test set. This was done for both multivariate and random forest regressions. Multivariate regression models were used to estimate

f_{p}

for RS-2 SAR scenes resampled to coarser 750 m by 750 m pixels to match the scale of the input sample data. To examine whether model estimates are sensitive to scale, images were also resampled to 500 m and 250 m pixel spacings.

2.3.5. Snow Thickness Images

Given the association of

f_{p}

and winter snow thickness [7,8], we infer the spatial distribution of relative snow thickness based on predicted

f_{p}

. To constrain the snow thickness, we assume that our in-situ-measured range of snow thickness (1.5 cm to 40 cm) is representative of the SAR scene extents (25 km by 25 km), which encompasses the study area. We linearly categorized the in-situ-measured snow thickness into eight classes, evenly-distributed across the snow thickness range, and related them to eight

f_{p}

classes ranging from 0.25–0.85 (i.e., lowest

f_{p}

class is equivalent to the highest class of snow thickness). This provides a rudimentary hindcast of snow thickness.

3. Results and Discussion

3.1. Relationship between Winter SAR Backscatter and Pond Fraction

Spearman rank correlation coefficients between RS-2 parameters and

f_{p}

are shown in Table 4.

σ_{H H}^{0}

is correlated with one or more parameters in each

θ

range and is thus not included in model development; it is shown in Table 4 for comparison. In the NR,

σ_{H V}^{0}

and

M E A N_{R_{V V H H}}

exceed 0.5, as does

σ_{H H}^{0}

. The value for

σ_{H V}^{0}

(−0.652) is within the range reported in Scharien et al. [24] (−0.647 to −0.792), but the value for

σ_{H H}^{0}

(−0.540) is slightly lower than their range (−0.590 to −0.824). Our relatively higher correlation value for

σ_{H V}^{0}

may relate to the lower noise floor of our data; suggesting that

σ_{H V}^{0}

was limited by the noise floor in Scharien et al. [24]. Our study examines data at a much lower noise floor: −37 dB (NR) to −33 dB (FR) for RS-2 fine quad-polarization mode (MDA, 2014), versus −20 dB to −22 dB for Envisat-ASAR and Sentinel-1 [68,69]. For

σ_{H V}^{0}

, our mean backscatter exceeds the RS-2 fine quad-polarization noise floor by 6.4 dB at NR, by 4.0 dB at MR, and by 3.1 dB at FR. In the MR, three parameters exceed 0.5:

M E A N_{H V}

M E A N_{H H}

and

A S M_{H V}

, plus

σ_{H H}^{0}

. Our correlation for

σ_{H H}^{0}

(−0.552) is lower than the range reported by Scharien et al. [24] (−0.695 to −0.810). However, our lower noise floor is likely the cause of our texture parameters derived from

σ_{H V}^{0}

exhibiting relatively strong correlation with

f_{p}

at MR when using RS-2, whereas the noise floor was found to be a limiting factor in Scharien et al. [24]. In the FR, three parameters exceed 0.5:

σ_{V V}^{0}

V A R_{V V}

and

M E A N_{A}

, plus

σ_{H H}^{0}

. We find

σ_{V V}^{0}

to consistently have higher correlation with

f_{p}

than

σ_{H H}^{0}

, suggesting that employing

σ_{V V}^{0}

instead of

σ_{H H}^{0}

may improve the results of Scharien et al. [24]. Overall, our correlation values are not as high as those found by Scharien et al. [24], which may be the result of their object-based approach versus our polygon-based approach. In comparison to Fors et al. [23], we find higher correlations for

H

α

ρ

and

ϕ

at FR, but lower correlations for

H

α

ρ

and

ϕ

at NR. Given that their X-band correlations are for sea ice with existing melt ponds, versus our C-band correlations for snow covered sea ice during winter, discrepancies are not unexpected.

The correlation analysis reveals multicollinearity between many parameters (r ≥ 0.7). Of the original 81 parameters, 64 are correlated at NR, 57 at MR and 62 at FR; the remaining less-correlated parameters are used for model development; these are shown in Table 4.

Given that late winter

σ_{V V}^{0}

and

σ_{H V}^{0}

exhibit negative correlation with melt season

f_{p}

(Table 4; e.g., Figure 4 and Figure 5), and that

f_{p}

is inversely correlated with late winter snow thickness [7], we infer that

σ_{V V}^{0}

, and

σ_{H V}^{0}

increase with snow thickness in a manner similar to previous studies [2,39].

R_{V V H H}

exhibits substantial negative correlation with

f_{p}

at MR (e.g., Figure 6).

R_{V V H H}

pixel values are primarily negative, indicating that

σ_{H H}^{0} > σ_{V V}^{0}

, suggesting that Fresnel reflection effects associated with a smooth dielectric interface are responsible [70].

R_{V V H H}

tends toward 0 dB as

f_{p}

decreases, pointing to reduced Fresnel reflection, which, in turn, may be associated with rougher FYI and thicker snow.

Texture parameters most correlated with

f_{p}

are primarily based on

σ_{H V}^{0},

σ_{V V}^{0}

σ_{H H}^{0}

R_{V V H H}

ϕ

and A (Table 4, Figure 7, Figure 8 and Figure 9). Texture parameters that are correlated with

f_{p}

and derived from

σ_{H V}^{0}

are present in all three

θ

ranges. Their correlation with

f_{p}

suggests that

σ_{H V}^{o}

might be sensitive to different snow thicknesses as well.

σ_{H V}^{o}

is influenced by surface roughness, volume scattering and multi-bounce scattering [71]; as snow thickness increases, the potential for increased volume scattering increases [72]. Thicker snow covers are associated with rougher sea ice [9], which may also influence

σ_{H V}^{o}

because cross polarization is the result of multiple volume scattering enhanced by the presence of uneven surfaces [73]. Therefore, it is not surprising that texture parameters associated with

σ_{H V}^{0}

are sensitive to ice and snow surface structure and related melt season

f_{p}

. Texture parameters based on

ϕ

and

A

are present in all three

θ

ranges, suggesting that these parameters may be sensitive to snow thickness differences and are therefore correlated with

f_{p}

3.2. Prediction of Pond Fraction

The regression analysis produced the following models for predicting

f_{p}

at NR, MR and FR, respectively:

f_{p (N R)} = 0.897 - 0.048 σ_{H V}^{0} + 0.726 E N T_{H H} - 0.059 M E A N_{ϕ} - 2.939 C O R_{α} + 0.020 α

(3)

f_{p (M R)} = 2.912 - 0.022 M E A N_{H V} - 0.052 M E A N_{ϕ} - 0.012 M E A N_{A} - 0.787 C O R_{H V} - 0.094 R_{V V H H}

(4)

f_{p (F R)} = 2.227 - 0.015 σ_{V V}^{0} - 0.059 M E A N_{ϕ} + 2.837 A S M_{H} - 0.015 M E A N_{A} - 0.174 V A R_{V V}

(5)

The models, as represented by Equations (3)–(5), exhibit

R^{2}

values of 0.57, 0.61, and 0.62 respectively. The FR model exhibits the highest

R^{2}

and contains a linear parameter and four texture parameters. The MR model exhibits the second highest

R^{2}

and contains one linear parameter and four texture parameters. All models make use of polarimetric data for the contributing texture parameters. The robustness of the models is consistent with previous findings in that the inclusion of texture with linear and/or polarimetric parameters improves classification results [48,49,50,51].

Differing scattering mechanisms at NR, MR and FR are very likely responsible for the three different models. Surface roughness changes depending on the local

θ

[74]; surface scattering dominates at NR, and volume scattering dominates at FR [41,43]. In addition, the polarimetric parameters that were found to have low correlation with other parameters are likely sensitive to specific scattering mechanisms; this may explain their inclusion in the models even if their correlation with

f_{p}

is low (Table 4). In comparison to the models developed by Scharien et al. [24], the model parameters here suggest that the addition of polarimetric capabilities can be a benefit, and a lower noise floor is a benefit when using HV. However, the use of additional model parameters may lead to over-fitting.

Random forest regression analysis provides predictive ability similar to the linear regression (see Section 3.3). An assessment of random forest parameter importance supports the parameter selection in the linear regression models (Table 5); most of the linear regression parameters are found in the top five important positions. Differences between models are expected due to their different selection techniques. Give the random forest results, no gross misspecification is anticipated in the linear models.

3.3. Error Analysis

Predicted

f_{p}

, using Equations (3)–(5), is compared with measured

f_{p}

for the 325 test samples, for each

θ

range (Figure 10). The RMSE was found to be 0.106 for NR, 0.108 for MR, and 0.094 for FR. The RMSE values obtained with random forest regression are 0.109, 0.091 and 0.092, for NR, MR and FR models, respectively. The NR linear model exhibits greater skill than the random forest model; however, the MR linear model is substantially worse that for random forest. Nevertheless, these values indicate that the multivariate linear regression models have predictive capabilities similar to that of random forest models and can thus be used without significant loss of skill.

It is apparent that the models overestimate lower

f_{p}

and underestimate higher

f_{p}

Figure 10. The underestimation of higher

f_{p}

is most pronounced at MR (Figure 10b). However, the linear dependencies are consistent across the three

θ

ranges.

3.4. SAR Scene Pond Fraction Estimates

The NR, MR and FR

f_{p}

prediction models are applied to late winter SAR scenes with pixel spacing of 750 m, 500 m and 250 m to assess the dependency of pixel size (Figure 11, Figure 12 and Figure 13). The areas represented by the models at lower and higher resolution indicate similar pattern and variability, indicating that the models are appropriate at higher resolution. The higher resolution SAR scene displays higher variability than the 750 m or 500 m scenes, which is logical as 250 m pixel spacing would reveal more detail. There are possible mixed-pixel effects where sea ice abuts land because the land mask used may not be perfectly accurate in all cases. However, these effects only impact the sea ice pixels immediately adjacent to the coastlines. Buffering the coastline would mitigate this issue, if needed.

Within the area of overlapping scenes (red outline in Figure 11a), some differences in the estimates can be observed. In general, the NR estimates (Figure 11) exhibit lower pond fractions than do the MR (Figure 12) and FR (Figure 13) estimates; especially at the north end of Griffith Island and associated with the linear feature extending from Sheringham Point. The latter is a region of rougher, ridged, sea ice which is expected to have thicker snow cover and smaller

f_{p}

. The efficacy of the NR model to show such smaller

f_{p}

is supported by the error analysis which shows that NR exhibits somewhat less bias at small

f_{p}

(Figure 10a). Conversely, the FR estimates show larger areas of higher

f_{p}

than MR and NR. This is also supported by the error analysis, which shows a smaller bias at high

f_{p}

for FR (Figure 10c). The MR estimates appear to lack the range of

f_{p}

, compared to the NR and FR estimates; this is also displayed in the error analysis (Figure 10b). Given these observations, it may be beneficial to make use of combined NR and FR estimates. The estimates of

f_{p}

with regard to incidence angle are very likely a function of the different scattering mechanisms that dominate at the respective incidence angle ranges and the sensitivity of the different parameter combinations in the models to these scattering mechanisms.

3.5. SAR-Based Estimation of Snow Thickness

The snow thickness of our in-situ sample site ranges from 1.5 cm to 40 cm, with a mean thickness of 10.9 cm ± 7.7 cm (Figure 2). This range of snow thickness is representative of FYI in the Canadian Arctic Archipelago; however, the mean snow thickness is lower than values previously found for relatively smooth FYI [75]. Considering the range of snow thickness in our study area, first-order hindcasts of snow thickness distribution were generated (Figure 14, Figure 15 and Figure 16) based on the association of snow thickness and

f_{p}

. The prediction maps provide insight into the snow thickness distribution within a 25 km by 25 km SAR scene. Of note is the higher accumulation of snow near the coastlines. This is because during wind drift, the uneven land surface topography causes more snow to accumulate on the leeward side [76] than in the homogeneous and smooth areas of FYI. The models illustrate a similar pattern of snow thickness distribution and show snow thickness variability at both fine and coarser resolution. Snow accumulation also depends on ice deformation [77], with snow catchment higher in deformed ice areas compared to smooth ice surfaces. The model hindcast maps indicating appreciably higher snow thickness along linear features near to the Cornwallis Island coastline appear to be reasonably well associated with the spatial location of narrow sea ice ridges as observed in the winter SAR imagery. Most of the study area exhibits a thin snow cover consistent with a relatively smooth surface. At both 750 m and 250 m pixel spacing, the models show similar pattern and variability of snow thickness.

Between the NR, MR and FR snow thickness estimates, NR (Figure 14) exhibits the greatest snow depths. This relates directly to the smaller

f_{p}

estimates with the NR model as discussed in the previous section. The reduced bias at small

f_{p}

with NR (Figure 10a) likely provides the best estimates of thick snow. Thinner snow estimates appear to have the most realistic spatial distribution at FR (Figure 16). Again, this relates directly to the smaller bias at large

f_{p}

with FR (Figure 10c), where thinner snow is expected.

4. Conclusions

This paper presents an approach utilizing late winter C-band RADARSAT-2 SAR linear, polarimetric and texture parameters of snow covered first-year sea ice to predict early summer melt pond fraction. Predictions were then used to examine relationships between early melt pond fraction and late winter snow thickness. We estimated melt pond fractions from the aerial photographs collected over first-year sea ice. The correlation of melt pond fraction and late winter backscatter parameters provided us with to the ability to hindcast snow thickness. Nine linear and polarimetric parameters and 72 texture parameters were evaluated for their relationship with pond fraction.

Multivariate models comprised of linear, polarimetric and/or texture parameters are derived at near-, mid- and far-range incidence angles. The best pond fraction prediction capability is exhibited by the model at far-range incidence angles (RMSE = 0.09).

By relating late winter snow thickness and

f_{p}

, these predictions help us to understand the winter snow thickness variability. The models are able to distinguish higher snow thickness along sea ice ridges, coastlines and relatively thinner snow cover on smooth surfaces of first-year sea ice, which is consistent with previous findings.

Moreover, the models show that the combination of SAR linear and polarimetric backscatter and texture parameters enhance the strength of the models compared to utilizing them separately for the prediction of pond fraction. The estimation of pond fraction over an entire SAR scene based on the models show logical distributions of melt ponds and snow thickness.

The results of this study add to the suite of seasonal sea ice forecasting tools, and thus can aid ship navigability since melt ponds are associated with advanced sea ice melt and significant weakening of sea ice mechanical strength. At the same time, it provides insight into the late winter snow thickness distribution on first year sea ice. The method is tested on landfast sea ice, is sensitive to the time periods of the collected aerial melt pond distributions and SAR scene acquisitions, and is likely not applicable to drift ice. Future research should test this model on a regional scale, and similar models should also be evaluated for their application over multi-year sea ice and more deformed types of first-year sea ice.

Author Contributions

S.R., T.G. and J.Y. designed the experiment. S.R. formulated the research methodology and wrote the manuscript. R.S. provided necessary data and contributed to manuscript development and proof reading. T.G. and J.Y. contributed with additional inputs during manuscript development and proof reading.

Funding

This research was funded by Canadian NSERC Discovery grants to John Yackel and Randy Scharien as well as MEOPAR (Marine Environmental Observation Prediction and Response Network) funding to Randy Scharien. The APC was funded by Canadian NSERC Discovery grants to John Yackel.

Acknowledgments

The authors would like to thank the participants of the Arctic-ICE 2012 Field Experiment based out of Resolute Bay, Nunavut, Canada. We would like to thank all team members for their support and hard work in the field program, including Principal Investigators, C.J. Mundy (CEOS, University of Manitoba) and B. Else (University of Calgary). We highly appreciate M.M. Rahman (University of Calgary) for his assistance in GLCM texture analysis. The authors would like to thank for the valuable comments provided by the reviewers; their contributions have substantially improved this paper. We also appreciate the collegial assistance from M. Mahmud and V. Nandan. We extend our sincere thanks to the Polar Continental Shelf Project and MEOPAR (Marine Environmental Observation Prediction and Response Network) for their funding and logistical support. We acknowledge Canadian NSERC Discovery grants to R. Scharien and J. Yackel towards this work. The Canadian Ice Service support in providing, planning and ordering RADARSAT-2 imagery is appreciated. Supplementary RADARSAT-2 imagery and planning support was provided by the Canadian Space Agency’s Science and Operational Applications Research program. RADARSAT-2 Data and Products © MacDonald, Dettwiler and Associates Ltd. 2012. All Rights Reserved. RADARSAT is an official mark of the Canadian Space Agency.

Conflicts of Interest

The authors declare no conflict of interest.

References

Holt, B.; Digby, S.A. Processes and imagery of first-year fast sea ice during the melt season. J. Geophys. Res. Oceans 1985, 90, 5045–5062. [Google Scholar] [CrossRef]
Barber, D.G.; Nghiem, S.V. The role of snow on the thermal dependence of microwave backscatter over sea ice. J. Geophys. Res. Oceans 1999, 104, 25789–25803. [Google Scholar] [Green Version]
Yackel, J.J.; Barber, D.G.; Hanesiak, J.M. Melt ponds on sea ice in the Canadian Archipelago. I–Variability in morphological and radiative properties. J. Geophys. Res. 2000, 105, 22049–22060. [Google Scholar] [CrossRef]
Scharien, R.K.; Yackel, J.J. Analysis of surface roughness and morphology of first-year sea ice melt ponds: Implications for microwave scattering. IEEE Trans. Geosci. Remote Sens. 2005, 43, 2927–2939. [Google Scholar] [CrossRef]
Iacozza, J.; Barber, D.G. Ablation patterns of snow cover over smooth first-year sea ice in the Canadian Arctic. Hydrol. Process. 2001, 15, 3559–3569. [Google Scholar] [CrossRef]
Eicken, H.; Krouse, H.R.; Kadko, D.; Perovich, D.K. Tracer studies of pathways and rates of meltwater transport through Arctic summer sea ice. J. Geophys. Res. Oceans 2002, 107. [Google Scholar] [CrossRef]
Yackel, J.J.; Barber, D.G.; Papakyriakou, T.N.; Breneman, C. First-year Sea ice spring melt transitions in the Canadian Arctic Archipelago from time-series synthetic aperture radar data, 1992–2002. Hydrol. Process. 2007, 21, 253–265. [Google Scholar] [CrossRef]
Polashenski, C.; Perovich, D.; Courville, Z. The mechanisms of sea ice melt pond formation and evolution. J. Geophys. Res. Oceans 2012, 117. [Google Scholar] [CrossRef] [Green Version]
Iacozza, J.; Barber, D.G. An examination of snow redistribution over smooth land-fast sea ice. Hydrol. Process. 2010, 24, 850–865. [Google Scholar] [CrossRef]
Iacozza, J.; Barber, D.G. An examination of the distribution of snow on sea-ice. Atmos. Ocean 1999, 37, 21–51. [Google Scholar]
Webster, M.A.; Rigor, I.G.; Perovich, D.K.; Richter-Menge, J.A.; Polashenski, C.M.; Light, B. Seasonal evolution of melt ponds on Arctic sea ice. J. Geophys. Res. Oceans 2015, 120, 5968–5982. [Google Scholar] [CrossRef] [Green Version]
Landy, J.; Ehn, J.; Shields, M.; Barber, D. Surface and melt pond evolution on landfast first-year sea ice in the Canadian Arctic Archipelago. J. Geophys. Res. Oceans 2014, 119, 3054–3075. [Google Scholar] [CrossRef] [Green Version]
Markus, T.; Cavalieri, D.J.; Tschudi, M.A.; Ivanoff, A. Comparison of aerial video and Landsat 7 data over ponded sea ice. Remote Sens. Environ. 2003, 86, 458–469. [Google Scholar] [CrossRef]
Tschudi, M.A.; Maslanik, J.A.; Perovich, D.K. Derivation of melt pond coverage on Arctic sea ice using MODIS observations. Remote Sens. Environ. 2008, 112, 2605–2614. [Google Scholar] [CrossRef]
Rösel, A.; Kaleschke, L.; Birnbaum, G. Melt ponds on Arctic sea ice determined from MODIS satellite data using an artificial neural network. Cryosphere 2012, 6, 431–446. [Google Scholar] [CrossRef] [Green Version]
Zege, E.; Malinka, A.; Katsev, I.; Prikhach, A.; Heygster, G.; Istomina, L.; Schwarz, P. Algorithm to retrieve the melt pond fraction and the spectral albedo of Arctic summer ice from satellite optical data. Remote Sens. Environ. 2015, 163, 153–164. [Google Scholar] [CrossRef] [Green Version]
Istomina, L.; Heygster, G.; Huntemann, M.; Schwarz, P.; Birnbaum, G.; Scharien, R.; Polashenski, C.; Perovich, D.; Zege, E.; Malinka, A.; et al. Melt pond fraction and spectral sea ice albedo retrieval from MERIS data-Part 1: Validation against in situ, aerial, and ship cruise data. Cryosphere 2015, 9, 1551–1566. [Google Scholar] [CrossRef] [Green Version]
Istomina, L.; Heygster, G.; Huntemann, M.; Marks, H.; Melsheimer, C.; Zege, E.; Malinka, A.; Prikhach, A.; Katsev, I. Melt pond fraction and spectral sea ice albedo retrieval from MERIS data–Part 2: Case studies and trends of sea ice albedo and melt ponds in the Arctic for years 2002–2011. Cryosphere 2015, 9, 1567–1578. [Google Scholar] [CrossRef]
Tanaka, Y.; Tateyama, K.; Kameda, T.; Hutchings, J.K. Estimation of melt pond fraction over high-concentration Arctic sea ice using AMSR-E passive microwave data. J. Geophys. Res. Oceans 2016, 121, 7056–7072. [Google Scholar] [CrossRef]
Barber, D.G.; Yackel, J. The physical, radiative and microwave scattering characteristics of melt ponds on Arctic landfast sea ice. Int. J. Remote Sens. 1999, 20, 2069–2090. [Google Scholar] [CrossRef]
Mäkynen, M.; Kern, S.; Rösel, A.; Pedersen, L.T. On the estimation of melt pond fraction on the Arctic Sea ice with ENVISAT WSM images. IEEE Trans. Geosci. Remote Sens. 2014, 52, 7366–7379. [Google Scholar] [CrossRef]
Scharien, R.K.; Landy, J.; Barber, D.G. First-year sea ice melt pond fraction estimation from dual-polarisation C-band SAR—Part 1: In situ observations. Cryosphere 2014, 8, 2147–2162. [Google Scholar] [CrossRef]
Fors, A.S.; Divine, D.V.; Doulgeris, A.P.; Renner, A.H.; Gerland, S. Signature of Arctic first-year ice melt pond fraction in X-band SAR imagery. Cryosphere 2017, 11, 755–771. [Google Scholar] [CrossRef] [Green Version]
Scharien, R.K.; Segal, R.; Nasonova, S.; Nandan, V.; Howell, S.E.; Haas, C. Winter Sentinel-1 backscatter as a predictor of spring Arctic sea ice melt pond fraction. Geophys. Res. Lett. 2017, 44. [Google Scholar] [CrossRef]
Scharien, R.; Segal, R.; Yackel, J.; Howell, S.; Nasonova, S. Linking winter and spring thermodynamic sea-ice states at critical scales using an object-based image analysis of Sentinel-1. Ann. Glaciol. 2017, 59, 1–15. [Google Scholar] [CrossRef]
Derksen, C.; Piwowar, J.; LeDrew, E. Sea-ice melt-pond fraction as determined from low level aerial photographs. Arct. Alp. Res. 1997, 29, 345–351. [Google Scholar] [CrossRef]
Tschudi, M.A.; Curry, J.A.; Maslanik, J.A. Determination of areal surface-feature coverage in the Beaufort Sea using aircraft video data. Ann. Glaciol. 1997, 25, 434–438. [Google Scholar] [CrossRef] [Green Version]
Barber, D.G.; Papakyriakou, T.N.; LeDrew, E.F. On the relationship between energy fluxes, dielectric properties, and microwave scattering over snow covered first-year sea ice during the spring transition period. J. Geophys. Res. Oceans 1994, 99, 22401–22411. [Google Scholar] [CrossRef]
Barber, D.G.; Reddan, S.P.; LeDrew, E.F. Statistical characterization of the geophysical and electrical properties of snow on landfast first-year sea ice. J. Geophys. Res. Oceans 1995, 100, 2673–2686. [Google Scholar] [CrossRef]
Sturm, M.; Holmgren, J.; Perovich, D.K. Winter snow cover on the sea ice of the Arctic Ocean at the Surface Heat Budget of the Arctic Ocean (SHEBA): Temporal evolution and spatial variability. J. Geophys. Res. Oceans 2002, 107. [Google Scholar] [CrossRef] [Green Version]
Bredow, J.W.; Gogineni, S. Comparison of measurements and theory for backscatter from bare and snow-covered saline ice. IEEE Trans. Geosci. Remote Sens. 1990, 28, 456–463. [Google Scholar] [CrossRef]
Lytle, V.I.; Jezek, K.C.; Hosseinmostafa, A.R.; Gogineni, S.P. Laboratory backscatter measurements over urea ice with a snow cover at Ku band. IEEE Trans. Geosci. Remote Sens. 1993, 31, 1009–1016. [Google Scholar] [CrossRef]
Markus, T.; Cavalieri, D.J. Snow depth distribution over sea ice in the Southern Ocean from satellite passive microwave data. In Antarctic Sea Ice: Physical Processes, Interactions and Variability; Wiley Online Library: Hoboken, NJ, USA, 1998; Volume 74, pp. 19–39. [Google Scholar]
Cavalieri, D.J.; Markus, T.; Ivanoff, A.; Miller, J.A.; Brucker, L.; Sturm, M.; Maslanik, J.A.; Heinrichs, J.F.; Gasiewski, A.J.; Leuschen, C.; et al. A comparison of snow depth on sea ice retrievals using airborne altimeters and an AMSR-E simulator. IEEE Trans. Geosci. Remote Sens. 2012, 50, 3027–3040. [Google Scholar] [CrossRef]
Maass, N.; Kaleschke, L.; Tian-Kunze, X.; Drusch, M. Snow thickness retrieval over thick Arctic sea ice using SMOS satellite data. Cryosphere 2013, 7, 1971–1989. [Google Scholar] [CrossRef] [Green Version]
Kurtz, N.T.; Farrell, S.L.; Studinger, M.; Galin, N.; Harbeck, J.P.; Lindsay, R.; Onana, V.D.; Panzer, B.; Sonntag, J.G. Sea ice thickness, freeboard, and snow depth products from Operation IceBridge airborne data. Cryosphere 2013, 7, 1035–1056. [Google Scholar] [CrossRef] [Green Version]
Kwok, R.; Kurtz, N.T.; Brucker, L.; Ivanoff, A.; Newman, T.; Farrell, S.L.; King, J.; Howell, S.; Webster, M.A.; Paden, J.; et al. Intercomparison of snow depth retrievals over Arctic sea ice from radar data acquired by Operation IceBridge. Cryosphere 2017, 11. [Google Scholar] [CrossRef]
Barber, D.G.; Fung, A.K.; Grenfell, T.C.; Nghiem, S.V.; Onstott, R.G.; Lytle, V.I.; Perovich, D.K.; Gow, A.J. The role of snow on microwave emission and scattering over first-year sea ice. IEEE Trans. Geosci. Remote Sens. 1998, 36, 1750–1763. [Google Scholar] [CrossRef]
Gill, J.P.; Yackel, J.J.; Geldsetzer, T.; Fuller, M.C. Sensitivity of C-band synthetic aperture radar polarimetric parameters to snow thickness over land fast smooth first-year sea ice. Remote Sens. Environ. 2015, 166, 34–49. [Google Scholar] [CrossRef]
Zheng, J.; Geldsetzer, T.; Yackel, J. Snow thickness estimation on first-year sea ice using microwave and optical remote sensing with melt modelling. Remote Sens. Environ. 2017, 199, 321–332. [Google Scholar] [CrossRef]
Kim, Y.S.; Onstott, R.; Moore, R. Effect of a snow cover on microwave backscatter from sea ice. IEEE J. Ocean. Eng. 1984, 9, 383–388. [Google Scholar]
Tjuatja, S.; Fung, A.K.; Bredow, J. A scattering model for snow-covered sea ice. IEEE Trans. Geosci. Remote Sens. 1992, 30, 804–810. [Google Scholar] [CrossRef]
Dierking, W.; Pettersson, M.I.; Askne, J. Multifrequency scatterometer measurements of Baltic Sea ice during EMAC-95. Int. J. Remote Sens. 1999, 20, 349–372. [Google Scholar] [CrossRef]
Wakabayashi, H.; Matsuoka, T.; Nakamura, K.; Nishio, F. Polarimetric characteristics of sea ice in the Sea of Okhotsk observed by airborne L-band SAR. IEEE Trans. Geosci. Remote Sens. 2004, 42, 2412–2425. [Google Scholar] [CrossRef]
Scheuchl, B.; Cumming, I.; Hajnsek, I. Classification of fully polarimetric single-and dual-frequency SAR data of sea ice using the Wishart statistics. Can. J. Remote Sens. 2005, 31, 61–72. [Google Scholar] [CrossRef]
Gill, J.P.; Yackel, J.J. Evaluation of C-band SAR polarimetric parameters for discrimination of first-year sea ice types. Can. J. Remote Sens. 2012, 38, 306–323. [Google Scholar] [CrossRef]
Gill, J.P.; Yackel, J.J.; Geldsetzer, T. Analysis of consistency in first-year sea ice classification potential of C-band SAR polarimetric parameters. Can. J. Remote Sens. 2013, 39, 101–117. [Google Scholar] [CrossRef]
Holmes, Q.A.; Nuesch, D.R.; Shuchman, R.A. Textural analysis and real-time classification of sea-ice types using digital SAR data. IEEE Trans. Geosci. Remote Sens. 1984, 22, 113–120. [Google Scholar] [CrossRef]
Shokr, M.E. Evaluation of second-order texture parameters for sea ice classification from radar images. J. Geophys. Res. Oceans 1991, 96, 10625–10640. [Google Scholar] [CrossRef]
Clausi, D.A. An analysis of co-occurrence texture statistics as a function of grey level quantization. Can. J. Remote Sens. 2002, 28, 45–62. [Google Scholar] [CrossRef]
Liu, H.; Guo, H.; Zhang, L. SVM-based sea ice classification using textural features and concentration from RADARSAT-2 Dual-Pol ScanSAR data. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2015, 8, 1601–1613. [Google Scholar] [CrossRef]
Haralick, R.M.; Shanmugam, K. Textural features for image classification. IEEE Syst. Man Cybern. Mag. 1973, 6, 610–621. [Google Scholar] [CrossRef]
Barber, D.; LeDrew, E. SAR sea ice discrimination using texture statistics—A multivariate approach. Photogramm. Eng. Remote Sens. 1991, 57, 385–395. [Google Scholar]
Soh, L.K.; Tsatsoulis, C. Texture analysis of SAR sea ice imagery using gray level co-occurrence matrices. IEEE Trans. Geosci. Remote Sens. 1999, 37, 780–795. [Google Scholar] [CrossRef] [Green Version]
Lee, J.S.; Wen, J.H.; Ainsworth, T.L.; Chen, K.S.; Chen, A.J. Improved sigma filter for speckle filtering of SAR imagery. IEEE Trans. Geosci. Remote Sens. 2009, 47, 202–213. [Google Scholar]
Oliver, C.; Quegan, S. Understanding Synthetic Aperture Radar Images; SciTech Publishing: Stevenage, UK, 2004. [Google Scholar]
Drinkwater, M.R.; Kwok, R.; Rignot, E.; Israelsson, H.; Onstott, R.G.; Winebrenner, D.P. Potential applications of polarimetry to the classification of sea ice. In Microwave Remote Sensing of Sea Ice; Carsey, F.D., Ed.; American Geophysical Union: Washington, DC, USA, 1992; pp. 419–428. [Google Scholar]
Cloude, S.R.; Pottier, E. An entropy based classification scheme for land applications of polarimetric SAR. IEEE Trans. Geosci. Remote Sens. 1997, 35, 68–78. [Google Scholar] [CrossRef]
Baraldi, A.; Parmiggiani, F. An investigation of the textural characteristics associated with gray level cooccurrence matrix statistical parameters. IEEE Trans. Geosci. Remote Sens. 1995, 33, 293–304. [Google Scholar] [CrossRef]
Hirose, T.K.; McNutt, L.; Paterson, J.S. A Study of Textoral and Tonal Information for Classifying Sea Ice from SAR Imagery. In 12th Canadian Symposium on Remote Sensing Geoscience and Remote Sensing Symposium, Proceedings of International Geoscience and Remote Sensing Symposium IGARSS 1989, Vancouver, BC, Canada, 10–14 July 1989; IEEE: New York, NY, USA, 1989; pp. 747–750. [Google Scholar]
Newsam, S.D.; Kamath, C. Retrieval Using Texture Features in High Resolution Multi-Spectral Satellite Imagery (No. UCRL-CONF-201981); Lawrence Livermore National Laboratory (LLNL): Livermore, CA, USA, 2004.
Gong, P.; Marceau, D.J.; Howarth, P.J. A comparison of spatial feature extraction algorithms for land-use classification with SPOT HRV data. Remote Sens. Environ. 1992, 40, 137–151. [Google Scholar] [CrossRef]
Cossu, R. Segmentation by means of textural analysis. Pixel 1988, 1, 21–24. [Google Scholar]
Liew, S.C.; Lim, H.; Kwoh, L.K.; Tay, G.K. Texture analysis of SAR images. In 1995 International Geoscience and Remote Sensing Symposium, IGARSS ’95. Quantitative Remote Sensing for Science and Applications, Proceedings of International Geoscience and Remote Sensing Symposium IGARSS 1995, Firenze, Italy, 10–14 July 1995; IEEE: New York, NY, USA, 1995; pp. 1412–1414. [Google Scholar]
Hall-Beyer, M. GLCM Texture: A Tutorial, Version 2.10. 2007. Available online: https://www.ucalgary.ca/mhallbey/tutorial-glcm-texture (accessed on 20 January 2018).
Breiman, L. Random Forests. Mach. Learn. 2001, 45, 5–32. [Google Scholar] [CrossRef] [Green Version]
Millard, K.; Richardson, M. On the importance of training data sample selection in random forest image classification: A case study in peatland ecosystem mapping. Remote Sens. 2015, 7, 8489–8515. [Google Scholar] [CrossRef]
Torres, R.; Buck, C.; Guijarro, J.; Suchail, J.L.; Schönenberg, A. The envisat ASAR instrument verification and characterisation. In Proceedings of the SAR workshop: CEOS Committee on Earth Observation Satellites, Toulouse, France, 26–29 October 1999; pp. 303–310. [Google Scholar]
Torres, R.; Snoeij, P.; Geudtner, D.; Bibby, D.; Davidson, M.; Attema, E.; Traver, I.N. GMES Sentinel-1 mission. Remote Sens. Environ. 2012, 120, 9–24. [Google Scholar] [CrossRef]
Nandan, V.; Scharien, R.; Geldsetzer, T.; Mahmud, M.; Yackel, J.J.; Islam, T.; Duguay, C. Geophysical and atmospheric controls on Ku-, X-and C-band backscatter evolution from a saline snow cover on first-year sea ice from late-winter to pre-early melt. Remote Sens. Environ. 2017, 198, 425–441. [Google Scholar] [CrossRef]
Scheuchl, B.; Flett, D.; Caves, R.; Cumming, I. Potential of RADARSAT-2 data for operational sea ice monitoring. Can. J. Remote Sens. 2004, 30, 448–461. [Google Scholar] [CrossRef]
Onstott, R.G. SAR and scatterometer signatures of sea ice. In Microwave Remote Sensing of Sea Ice; Carsey, F.D., Ed.; American Geophysical Union: Washington, DC, USA, 1992; pp. 73–104. [Google Scholar]
Fung, A.K.; Eom, H.J. Application of a combined rough surface and volume scattering theory to sea ice and snow backscatter. IEEE Trans. Geosci. Remote Sens. 1982, 4, 528–536. [Google Scholar] [CrossRef]
Lewis, A.J.; Henderson, F.M.; Holcomb, D.W. Radar fundamentals: the geoscience perspective. In Principles & Applications of Imaging Radar. Manual of Remote Sensing, 3rd ed.; Wiley: Hoboken, NJ, USA, 1998; Volume 3, pp. 131–180. [Google Scholar]
Nandan, V.; Geldsetzer, T.; Yackel, J.; Mahmud, M.; Scharien, R.; Howell, S.; King, J.; Ricker, R.; Else, B. Effect of Snow Salinity on CryoSat-2 Arctic First- Year Sea Ice Freeboard Measurements. Geophys. Res. Lett. 2017, 44. [Google Scholar] [CrossRef]
Sturm, M.; Liston, G.E.; Benson, C.S.; Holmgren, J. Characteristics and growth of a snowdrift in Arctic Alaska, USA. Arct. Antarct. Alp. Res. 2001, 33, 319–329. [Google Scholar] [CrossRef]
Dierking, W.; Dall, J. Sea-ice deformation state from synthetic aperture radar imagery—Part I: Comparison of C-and L-band and different polarization. IEEE Trans. Geosci. Remote Sens. 2007, 45, 3610–3622. [Google Scholar] [CrossRef]

Figure 1. Study area in Resolute Passage, Nunavut, illustrating the in-situ field site, the flightpath of the aerial survey, and the region covered by RADARSAT-2 scenes. The red box in the inset map of the Canadian Arctic Archipelago delineates the location of the study area.

Figure 2. (a) Distribution of 202 in-situ snow thickness measurements collected on FYI acquired from the field site during Leg 1. (b) Hourly air temperature at the Resolute CARS weather station, Nunavut (blue), and once a minute on-ice air temperature (green).

Figure 3. (a) Sample of a classified aerial photograph with melt ponds (grey areas) and snow patches (white areas); (b) histogram of pond fraction distribution for the flightpath on 22 June 2012.

Figure 4. Late winter

σ_{H V}^{0} (dB)

as a function of

f_{p}

at NR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 4. Late winter

σ_{H V}^{0} (dB)

as a function of

f_{p}

at NR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 5. Late winter

σ_{V V}^{0} (dB)

as a function of

f_{p}

at FR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 5. Late winter

σ_{V V}^{0} (dB)

as a function of

f_{p}

at FR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 6. Late winter

R_{V V H H} (dB)

as a function of

f_{p}

at MR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 6. Late winter

R_{V V H H} (dB)

as a function of

f_{p}

at MR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 7. Late winter

M E A N_{H V}

, as a function of

f_{p}

at MR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 7. Late winter

M E A N_{H V}

, as a function of

f_{p}

at MR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 8. Late winter

M E A N_{ϕ}

, as a function of

f_{p}

at FR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 8. Late winter

M E A N_{ϕ}

, as a function of

f_{p}

at FR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 9. Late winter

C O R_{A}

, as a function of

f_{p}

at NR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 9. Late winter

C O R_{A}

, as a function of

f_{p}

at NR. The solid line represents the linear dependency, and r is the Spearman correlation.

Figure 10. Observed (estimated via aerial photography) and regression model predicted

f_{p}

at (a) NR, (b) MR and (c) FR. The dashed line is the linear dependency of the comparison. The solid line represents the 1:1 relationship.

Figure 10. Observed (estimated via aerial photography) and regression model predicted

f_{p}

at (a) NR, (b) MR and (c) FR. The dashed line is the linear dependency of the comparison. The solid line represents the 1:1 relationship.

Figure 11. Predicted melt

f_{p}

at 750 m (a), 500 m (b) and 250 m (c) pixel spacing, respectively, at NR. Based on RS-2 FQ4 2012-05-10 data. The red outline in (a) delineates the area where the NR, MR and FR scenes overlap.

Figure 11. Predicted melt

f_{p}

at 750 m (a), 500 m (b) and 250 m (c) pixel spacing, respectively, at NR. Based on RS-2 FQ4 2012-05-10 data. The red outline in (a) delineates the area where the NR, MR and FR scenes overlap.

Figure 12. Predicted melt

f_{p}

at 750 m (a), 500 m (b) and 250 m (c) pixel spacing, respectively, at MR. Based on RS-2 FQ15 2012-05-21 data.

Figure 12. Predicted melt

f_{p}

at 750 m (a), 500 m (b) and 250 m (c) pixel spacing, respectively, at MR. Based on RS-2 FQ15 2012-05-21 data.

Figure 13. Predicted melt

f_{p}

at 750 m (a), 500 m (b) and 250 m (c) pixel spacing, respectively, at FR. Based on RS-2 FQ23 2012-05-15 data.

Figure 13. Predicted melt

f_{p}

at 750 m (a), 500 m (b) and 250 m (c) pixel spacing, respectively, at FR. Based on RS-2 FQ23 2012-05-15 data.

Figure 14. Snow thickness class spatial distribution from the NR model at 750 m (a) and 250 m (b) pixel spacing. Grey area represents land. Based on RS-2 FQ4 2012-05-10 data.

Figure 15. Snow thickness class spatial distribution from the MR model at 750 m (a), and 250 m (b) pixel spacing, respectively. Grey area represents land. Based on RS-2 FQ15 2012-05-21 data.

Figure 16. Snow thickness class spatial distribution from the FR model at 750 m (a) and 250 m (b) pixel spacing. Grey area represents land. Based on RS-2 FQ23 2012-05-15 data.

Table 1. 25 km by 25 km RS-2 scenes used in this study and the number of 750 m by 750 m aerial photographs that fall within each SAR scene.

SAR Scene	Acquisition Date (yyyy-mm-dd)	Incidence Angle (Range)	Number of Aerial Photographs
FQ4	2012-05-10	23.1° (NR)	78
FQ8	2012-05-24	27.8° (NR)	97
FQ13	2012-05-21	33.2° (MR)	97
FQ15	2012-05-21	35.2° (MR)	87
FQ18	2012-05-18	38.1° (MR)	102
FQ22	2012-05-17	41.7° (FR)	87
FQ23	2012-05-15	42.6° (FR)	102

Table 2. Parameters employed in this study. Matrix elements are obtained from a standard covariance matrix (1) and coherency matrix (2) (Adapted from Gill et al., 2015).

Nomenclature	Notation	Reference
Co-polarized power (horizontal) (in dB)	$σ_{H H}^{0}$ (linear)	[57]
Co-polarized power (vertical) (in dB)	$σ_{V V}^{0}$ (linear)
Cross-polarized power (in dB)	$σ_{H V}^{0}$ (linear)
Co-polarized ratio (in dB)	$R_{V V H H} = \frac{〈 S_{V V} S_{V V}^{} 〉}{〈 S_{H H} S_{H H}^{} 〉}$ (linear)
Co-polarized phase difference	$ϕ = \tan^{- 1} [\frac{I m 〈 S_{H H} S_{V V}^{} 〉}{R e 〈 S_{H H} S_{V V}^{} 〉}]$ (polarimetric)
Co-polarized correlation co-efficient	$ρ = \| \frac{〈 S_{H H} S_{V V}^{} 〉}{\sqrt{〈 S_{H H} S_{H H}^{} S_{V V} S_{V V}^{*} 〉}} \|$ (polarimetric)
Entropy	$H$ (polarimetric)	[58]
Anisotropy	$A$ (polarimetric)
Alpha angle	$α$ (polarimetric)

Table 3. GLCM texture parameters used in the study.

Terminology	Theoretical Description	References
Entropy (ENT)	Detects the randomness of grey level distribution. Entropy will be higher for nearly random or noisy images.	[59]
Angular second moment (ASM)	Detects disorder in texture. High energy values represent the constant of periodic form of gray level distribution	[53,60]
Contrast (CON)	The difference between the highest and the lowest values of a connecting set of pixels. Contrast is greater for images that have rapidly fluctuating intensities.	[61]
GLCM variance (VAR)	The degree of heterogeneity. Variance increases when the gray level values differ from their mean.	[62]
Correlation (COR)	Measures linear-dependencies of gray tone in the image.	[63]
Homogeneity (HOM)	Gives maximum value when all elements in the image are same.	[49]
Dissimilarity (DIS)	Assists to distinguish most surface features.	[64]
GLCM Mean (MEAN)	Measures the mean of Grey level co-occurrences values. Generally least correlated with other texture parameters.	[65]

Table 4. Parameters used for model development. Parameters have low correlation with each other (Spearman’s r < 0.7). Parameters are shown in descending order of absolute correlation (Spearman’s) with

f_{p}

. NR = near-range, MR = mid-range, and FR = far-range incidence angles.

σ_{H H}^{0}

(value in italics) is included for comparison.

f_{p}

. NR = near-range, MR = mid-range, and FR = far-range incidence angles.

σ_{H H}^{0}

(value in italics) is included for comparison.

NR Parameters	r	MR Parameters	r	FR Parameters	r
$σ_{H V}^{0}$	−0.652	$M E A N_{H V}$	−0.627	$σ_{V V}^{0}$	−0.584
$σ_{H H}^{0}$	−0.540	$M E A N_{H H}$	−0.579	$M E A N_{A}$	−0.550
$M E A N_{R_{V V H H}}$	−0.532	$σ_{H H}^{0}$	−0.552	$σ_{H H}^{0}$	−0.540
$C O R_{A}$	0.473	$A S M_{H V}$	0.500	$V A R_{V V}$	−0.501
$C O R_{R_{V V H H}}$	−0.336	$R_{V V H H}$	−0.497	$M E A N_{ϕ}$	−0.464
$H O M_{H H}$	−0.323	$C O R_{H V}$	−0.427	$H O M_{H V}$	0.440
$E N T_{H H}$	0.304	$V A R_{V V}$	−0.361	$H$	0.382
$M E A N_{V V}$	−0.285	$M E A N_{ϕ}$	−0.351	$H O M_{V V}$	0.364
$V A R_{R_{V V H H}}$	−0.271	$M E A N_{R_{V V H H}}$	−0.344	$R_{V V H H}$	−0.312
$C O R_{V V}$	−0.253	$A$	−0.299	$A S M_{ρ}$	−0.293
$C O R_{α}$	−0.224	$ϕ$	−0.285	$H O M_{α}$	−0.211
$C O R_{ϕ}$	−0.196	$V A R_{H H}$	−0.259	$C O R_{A}$	−0.156
$C O R_{ρ}$	−0.177	$M E A N_{A}$	−0.242	$A S M_{A}$	−0.116
$H O M_{H V}$	0.164	$V A R_{A}$	−0.241	$V A R_{A}$	−0.105
$H O M_{ρ}$	−0.135	$V A R_{ϕ}$	−0.229	$H O M_{H}$	−0.072
$M E A N_{ϕ}$	0.080	$C O R_{α}$	−0.206	$D I S_{H}$	0.066
$α$	0.065	$C O R_{H H}$	−0.180	$A S M_{H}$	−0.058
		$V A R_{R_{V V H H}}$	−0.164	$C O N_{A}$	−0.052
		$A S M_{ϕ}$	0.154	$H O M_{A}$	−0.04
		$A S M_{V V}$	0.152
		$A S M_{R_{V V H H}}$	0.113
		$C O R_{ϕ}$	−0.105
		$A S M_{H H}$	0.073
		$α$	−0.050
		$ρ$	0.042

Table 5. Random forest regression parameters in order of parameter importance (increase in mean-squared-error with removal (IncMSE)), for NR, MR and FR models. The top 13 important positions are shown. Shaded parameters are found in the respective linear regression models.

NR	IncMSE	MR	IncMSE	FR	IncMSE
$σ_{H V}^{0}$	0.0116	$M E A N_{H V}$	0.0058	$M E A N_{ϕ}$	0.0046
$M E A N_{R_{V V H H}}$	0.0036	$A S M_{H V}$	0.0038	$M E A N_{A}$	0.0039
$M E A N_{V V}$	0.0015	$M E A N_{H H}$	0.0036	$σ_{V V}^{0}$	0.0035
$E N T_{H H}$	0.0010	$R_{V V H H}$	0.0029	$H O M_{V V}$	0.0021
$C O R_{α}$	0.0010	$M E A N_{A}$	0.0021	$V A R_{V V}$	0.0018
$H O M_{H H}$	0.0009	$A$	0.0021	$H O M_{H V}$	0.0017
$C O R_{A}$	0.0009	$M E A N_{ϕ}$	0.0017	$R_{V V H H}$	0.0008
$α$	0.0005	$ϕ$	0.0013	$C O R_{A}$	0.0007
$C O R_{V V}$	0.0004	$C O R_{H V}$	0.0012	$H$	0.0006
$V A R_{R_{V V H H}}$	0.0004	$V A R_{A}$	0.0008	$A S M_{ρ}$	0.0003
$M E A N_{ϕ}$	0.0004	$V A R_{V V}$	0.0007	$A S M_{H}$	0.0002
$H O M_{ρ}$	0.0003	$V A R_{R_{V V H H}}$	0.0006	$A S M_{A}$	0.0002
$H O M_{H V}$	0.0002	$A S M_{R_{V V H H}}$	0.0005	$H O M_{A}$	0.0002

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Ramjan, S.; Geldsetzer, T.; Scharien, R.; Yackel, J. Predicting Melt Pond Fraction on Landfast Snow Covered First Year Sea Ice from Winter C-Band SAR Backscatter Utilizing Linear, Polarimetric and Texture Parameters. Remote Sens. 2018, 10, 1603. https://doi.org/10.3390/rs10101603

AMA Style

Ramjan S, Geldsetzer T, Scharien R, Yackel J. Predicting Melt Pond Fraction on Landfast Snow Covered First Year Sea Ice from Winter C-Band SAR Backscatter Utilizing Linear, Polarimetric and Texture Parameters. Remote Sensing. 2018; 10(10):1603. https://doi.org/10.3390/rs10101603

Chicago/Turabian Style

Ramjan, Saroat, Torsten Geldsetzer, Randall Scharien, and John Yackel. 2018. "Predicting Melt Pond Fraction on Landfast Snow Covered First Year Sea Ice from Winter C-Band SAR Backscatter Utilizing Linear, Polarimetric and Texture Parameters" Remote Sensing 10, no. 10: 1603. https://doi.org/10.3390/rs10101603

APA Style

Ramjan, S., Geldsetzer, T., Scharien, R., & Yackel, J. (2018). Predicting Melt Pond Fraction on Landfast Snow Covered First Year Sea Ice from Winter C-Band SAR Backscatter Utilizing Linear, Polarimetric and Texture Parameters. Remote Sensing, 10(10), 1603. https://doi.org/10.3390/rs10101603

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

Predicting Melt Pond Fraction on Landfast Snow Covered First Year Sea Ice from Winter C-Band SAR Backscatter Utilizing Linear, Polarimetric and Texture Parameters

Abstract

1. Introduction

Objectives

2. Methodology

2.1. Study Area

2.2. Data

2.2.1. Aerial Surveys

2.2.2. RADARSAT-2 Data

2.3. Methods

2.3.1. Estimation of Melt Pond Fraction from Aerial Photographs

2.3.2. RADARSAT-2 Data Processing

2.3.3. Image Sampling

2.3.4. Regression Model Development

2.3.5. Snow Thickness Images

3. Results and Discussion

3.1. Relationship between Winter SAR Backscatter and Pond Fraction

3.2. Prediction of Pond Fraction

3.3. Error Analysis

3.4. SAR Scene Pond Fraction Estimates

3.5. SAR-Based Estimation of Snow Thickness

4. Conclusions

Author Contributions

Funding

Acknowledgments

Conflicts of Interest

References

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI